Board Members’ Overseas Experience and Foreign Investors’ Holdings: Evidence from China’s A-Share Market

Abstract

We examine the effect of directors’ overseas experience on foreign investors’ holdings using a large panel of Chinese listed firms from 2009 to 2022. We find that firms with a higher proportion of overseas-experienced directors exhibit significantly greater foreign institutional ownership. This positive association is robust to alternative variable definitions, model specifications, and sample restrictions, and it remains after addressing endogeneity concerns. Further analysis shows that the effect is stronger when overseas-experienced directors are more likely to function as effective monitors and advisors. Overall, our findings suggest that overseas experience helps transmit governance practices, knowledge, and skills valued by global investors and provides a credible signal that attracts foreign capital.

1. Introduction

Foreign capital has emerged as a critical source of financing for both developed and emerging economies. As one of the world’s fastest-growing economies, China has progressively liberalized its financial markets over the past two decades to attract global investors. Foreign institutional participation in China’s onshore A-share market has expanded mainly through three channels: the Qualified Foreign Institutional Investor (QFII) program launched in 2002, the Renminbi QFII (RQFII) program introduced in 2011, and the Hong Kong Stock Connect (HKSC) programs initiated in 2014 (see Lundblad et al., 2025 for a detailed discussion of these channels). The inclusion of China A-shares in the MSCI Emerging Markets Index in 2018 marked a significant milestone, substantially expanding global investors’ access to China’s capital market. By the end of 2021, foreign ownership accounted for approximately 5% of the aggregate A-share market capitalization (Lundblad et al., 2025). This share is likely to keep rising as Chinese regulators continue to lower barriers to participation and broaden market access, such as the 2020 removal of quota limits for QFII and RQFII. Given the growing role of foreign institutions in China’s equity market, identifying the factors that attract foreign capital is increasingly important.

Corporate boards play a crucial role in shaping firms’ governance, strategic decisions, and performance, factors that are central to foreign institutional investors’ asset-allocation choices. In this paper, we contribute to the literature by examining whether a salient dimension of board expertise—overseas experience—affects foreign institutional ownership in China, the world’s second-largest equity market. This question is particularly timely given the rapid growth in the number of Chinese professionals with educational and/or professional experience in developed markets such as the U.S. and U.K., alongside a significant increase in returnee talent in recent years (Yuan & Wen, 2018).

When investing in China’s A-share market, foreign investors encounter substantial frictions stemming from differences in language and culture, accounting and disclosure practices, and the legal and institutional environment (e.g., Covrig et al., 2007; Chen et al., 2020). These frictions exacerbate information asymmetry and raise monitoring costs, making it more challenging for foreign institutions to identify and invest in high-quality firms. We posit that the presence of directors with overseas educational and/or work experience serves as a mechanism to mitigate these challenges through three specific channels.

First, overseas-experienced directors are likely to foster better corporate governance (Giannetti et al., 2015; Tao et al., 2022). Exposure to markets with mature legal systems and strong investor protection helps these directors internalize higher standards regarding board oversight, internal controls, and accountability. In addition, compared to purely local directors, returnee directors may be less embedded in local political and business networks, making them less susceptible to pressure from local constituencies and more independent in their monitoring role. To the extent that foreign institutions prefer better-governed firms, they may allocate more capital to firms with overseas-experienced directors (Leuz et al., 2009). Second, overseas experience constitutes a valuable form of human capital that enhances firm performance. Directors with international exposure possess specialized knowledge, managerial expertise, and global networks that are relatively scarce in the domestic labor market. This human capital facilitates key strategic initiatives such as cross-border acquisitions, international capital raising, export expansion, and corporate innovation, which ultimately drive productivity and firm value (e.g., Bloom & Van Reenen, 2007; Filatotchev et al., 2009; Giannetti et al., 2015; Yuan & Wen, 2018). These improvements make the firm more attractive to foreign institutional investors. Third, overseas-experienced directors help bridge the information gap between the firm and foreign investors. Their familiarity with foreign languages, international business norms, and global disclosure standards facilitates communication with international investors and enhances the credibility and accessibility of firm information. This, in turn, reduces foreign investors’ information acquisition and monitoring costs, thereby encouraging foreign investment (DeFond et al., 2011).

Conversely, it is plausible that overseas experience has a negligible or even negative effect on foreign ownership. Directors with overseas backgrounds might leverage their informational and communication advantages to extract private benefits rather than serve shareholders. If their specialized expertise is difficult for other board members and outside investors to evaluate, they may negotiate terms in complex transactions, such as cross-border acquisitions, that disproportionately benefit themselves (Liu & Kang, 2025). Furthermore, in emerging markets with weaker external investor protection, controlling shareholders and management may constrain board independence, limiting directors’ ability to improve governance in practice (Liao et al., 2022). In such cases, overseas experience may be largely symbolic and fail to change firm behavior in ways that matter to foreign institutions. Taken together, these competing predictions imply that the net effect of board overseas experience on foreign institutional ownership is ultimately an empirical question.

To empirically test the conjectures above, we construct a large panel of Chinese listed firms using data from the China Stock Market and Accounting Research (CSMAR) Database and the Huazheng Rating Database over the period 2009–2022. Our main results show that the share of directors with overseas experience is positively associated with foreign institutional ownership. This relation is both statistically and economically significant. Specifically, a one standard deviation increase in the proportion of overseas-experienced directors is associated with a 17.87% increase relative to the sample mean of foreign institutional ownership. We conduct a battery of robustness tests using alternative variable definitions and sample restrictions, and our inferences remain unchanged. Further analyses indicate that the positive relation is stronger when overseas-experienced directors are more likely to be effective monitors and advisors, proxied by superior environmental, social, and governance (ESG) performance.

Our baseline results may be subject to two endogeneity concerns. First, omitted-variable bias may arise if unobserved firm characteristics jointly affect both the appointment of overseas-experienced directors and foreign institutional holdings. Second, reverse causality is possible if foreign investors prefer firms that already have internationally experienced directors, or if their presence influences subsequent board composition. To mitigate these concerns, we employ an instrumental-variable approach. Specifically, we use the one-year lagged proportion of non-mainland-China-born directors as an instrument for the share of directors with overseas experience. The results remain robust, supporting a causal interpretation of the effect of board overseas experience on foreign institutional ownership, although endogeneity concerns cannot be ruled out entirely.

Our study contributes to the extant literature in two ways. First, we extend the growing body of research on the economic consequences of board human capital, with a particular focus on directors’ overseas experience. Prior studies have largely examined how returnee directors influence corporate policies, such as disclosure quality (Liao et al., 2022), dividend payout (Tao et al., 2022), export expansion (Filatotchev et al., 2009), innovation (Yuan & Wen, 2018), corporate social responsibility (Zhang et al., 2018), and the cost of debt (Khan et al., 2023). We instead focus on the capital market consequence of overseas board expertise: foreign institutional ownership. This setting is particularly relevant because foreign institutions have become an increasingly important shareholder group in China’s equity market. By documenting a positive association between directors’ overseas experience and foreign ownership, our evidence complements existing studies and suggests that overseas experience facilitates the transmission of governance practices, knowledge, and skills valued by global investors.

Second, our findings enrich the literature on the determinants of foreign equity ownership by identifying board overseas experience as an important firm-level attribute that attracts foreign investors. Existing work emphasizes country- and firm-level frictions such as financial reporting and accounting standards (DeFond et al., 2011), levels of investor protection (Aggarwal et al., 2005; Leuz et al., 2009), social trust (Jin et al., 2016), cross-listing on U.S. exchanges (Ammer et al., 2012), and board structure (Miletkov et al., 2014). We add to this literature by highlighting a largely self-determined governance feature: appointing directors with overseas experience. This mechanism is particularly salient in China, where the growing pool of internationally trained professionals and returning talent is increasingly reshaping the governance practices of listed firms.

The remainder of this paper is structured as follows. Section 2 reviews the related literature. Section 3 describes data and methodology. Section 4 presents the main results and robustness analyses. Section 5 concludes.

2. Literature Review

By examining the effect of directors’ overseas experience on foreign investors’ holdings, we bring together two different strands of the literature.

2.1. Foreign Investors’ Holdings

Foreign investors’ holdings, also referred to as foreign institutional ownership from the firms’ perspective, represent equity stakes in Chinese A-share firms held by non-domestic institutional investors. In our setting, it specifically includes holdings by Qualified Foreign Institutional Investors (QFIIs), a scheme introduced by Chinese regulators to grant select foreign institutions access to the A-share market (Dahlquist & Robertsson, 2001; Jin et al., 2016). (The Qualified Foreign Institutional Investor (QFII) program, introduced by Chinese regulators in 2002, allows approved foreign institutional investors to invest directly in RMB-denominated A-shares listed on China’s mainland stock exchanges. Initially subject to licensing and investment quotas, the program was gradually expanded over time, with quotas fully removed in 2019 and the QFII and RQFII schemes unified in 2020, substantially broadening foreign investors’ access to China’s domestic capital markets.) Most of the work being done in the field of international finance comes to the same conclusion: international investors place their investments in areas where the information is comparable, governance is reliable, and frictions are low. Dahlquist and Robertsson (2001) use exhaustive ownership data and show that firm size, market liquidity, and an analyzable information environment strongly correlate with foreign holdings, indicating that informatively “legible” companies are easier for outsiders to own. With this same spirit, DeFond et al. (2011) show that mandatory IFRS adoption tends to attract more foreign mutual fund investment because it makes firms’ financial reports easier to compare and reduces the effort required to analyze and benchmark their statements. The related reduction in home bias through internationally familiar reporting is also well documented; for example, Covrig et al. (2007) demonstrate that early IAS adoption is associated with increased foreign mutual fund holdings, as reporting became more globally intelligible.

Compositional aspects of institutional capital are equally important. Ferreira and Matos (2008) discuss the “colors of money.” Different institutions and types of investors have varying time and attention spans and monitoring levels which affect governance and ownership differently across countries. Similarly, Aggarwal et al. (2005) report that foreign institutions focus on firms that are relatively easier to value and manage. However, Bena et al. (2017) take a longer view and focus on changes: foreigners are not just passing “locusts.” Controlling momentum and fundamental variables, increases in foreign institutional ownership are linked to enhanced governance quality and value gains that endure over time.

Moreover, foreign institutional ownership can also be influenced by legal and macro-institutional factors for the emerging markets. Batten and Vo (2015) consider the case that stronger legal protection, deeper markets, and macro stability promote foreign participation. On the disclosure side, Abdioglu et al. (2015) argue that greater transparency and mandatory regulatory reporting reduce information asymmetry and attract foreign institutions. At the same time, soft-information environments also play a role in this setting. Jin et al. (2016) show that greater social trust in China is associated with larger QFII holdings, indicating that credible social and institutional environments can lower perceived risks for cross-border investors.

Finally, the friction in the foreign holding may come from the language and access barriers, which can influence foreign institutional ownership. Korkeamäki et al. (2019) document that as foreign investors “learn Chinese,” their behavior in China’s equity market shifts in ways consistent with diminishing information barriers. Closely related, Bao and Wei (2024) show that sustainability reports issued in multiple languages raise foreign investors’ willingness to hold the stock, reinforcing the broader conclusion that when firms become legible to a global audience, foreign ownership tends to follow. Taken together, these studies indicate that comparability like IFRS, transparent disclosure, robust governance, and accessible communication are central screens through which foreign institutions decide whether to enter and remain invested or the influencing factors.

2.2. Overseas Experience of Board Members

Overseas experience of board members (BOE) refers to directors’ prior education or professional employment abroad; empirically, we operationalize BOE as the share of board members with foreign experience, an attribute repeatedly linked to stronger monitoring, lower opacity, and more conservative compliance in Chinese firms (Cao et al., 2019; Wen et al., 2020; Xiang et al., 2024). Against this preference map, the BOE functions as a composite signal, one that sophisticated global investors can verify and price. On the monitoring and information channel, BOE is associated with fewer opaque episodes and more credible oversight. For Chinese A-share companies, stock-price crash risk is typically lower in firms where board members possess international experience, implying less bad-news hoarding and a cleaner information environment (Cao et al., 2019). On the compliance margin, Wen et al. (2020) show that lower corporate tax avoidance is typically observed in firms led by managers with foreign experience, which complements the notion that BOE fosters a conservative, rule-compatible posture that outside investors prefer. In a complementary governance frame, Xiang et al. (2024) find that foreign-experienced directors are more effective monitors, improving investment efficiency, again aligning with the governance screen prominent in the foreign ownership literature.

On the resource and strategy channel, meanwhile, BOE embeds cross-border networks and knowledge that can enhance long-horizon value. Board internationalization tends to co-move with stronger CSR practices (Luo et al., 2022), thereby reducing reputational risk and strengthening stakeholder alignment. At the executive level, CEO education and foreign experience are associated with innovation outcomes and environmental innovation (Xiang et al., 2024; Zhou et al., 2021). In the Chinese context, foreign-experienced executives adjust capital structures toward targets more effectively (Gu et al., 2023), and managerial foreign experience predicts outward FDI, which is an indicator of strategic capability that international owners are likely to reward (Wen et al., 2023). These traits are observable to external investors and broadly consistent with the way foreign institutions screen for quality (Ferreira & Matos, 2008).

Equally important, a practical legibility mechanism amplifies these effects, where governance improvements like BOE are embedded within globally readable disclosure ecosystems, and outsiders can more readily validate and price them. The finding that multilingual sustainability reporting increases foreign ownership (see, e.g., Bao & Wei, 2024) dovetails with this mechanism, which sits naturally alongside evidence that outsiders respond to comparable, well-explained information (Covrig et al., 2007; DeFond et al., 2011).

3. Data and Methodology

3.1. Data

Our principal data repositories are from the CSMAR Database (China Stock Market & Accounting Research Database) and the Huazheng (CSI) ESG Database. We restrict the scope of our empirical work to firms listed on China’s A-share market. Given that overseas experience of board members is a low-frequency, relatively stable board attribute typically disclosed in annual reports, we use annual data in our analysis. To keep all variables on the same cadence and match the availability of ESG information, we select a sample period spanning from 2009 to 2022. Furthermore, to maintain comparability of foreign ownership measures and accounting ratios, we exclude firms in the financial sector as well as ST and *ST firms, whose leverage regimes and trading or reporting constraints differ materially from other industries.

Our independent variable is the overseas experience of board members (BOE), which is constructed using CSMAR’s Directors, Supervisors and Senior Executives Characteristics datasets. In our study, we follow Liu and Kang (2025) and code a board member as one with overseas experience if their résumé indicates overseas education and/or employment. The value is 1 if a board member has overseas education or work experience, and 0 otherwise. Next, we construct the board-level BOE measure as the count of overseas-experienced board members divided by board size. As a robustness check, we generate two new related variables. First, the high-overseas dummy variable (BOE Dummy) is constructed by first computing the median value of BOE across all firms in each year and then assigning a value of 1 if a firm’s BOE in the following year exceeds that annual median, and zero otherwise. Second, we use the natural logarithm of BOE plus one as another alternative measure of the independent variable, denoted as LnBOE.

From CSMAR, we also obtain our dependent variable, Foreign Investors’ Holdings, proxied by the QFII (Qualified Foreign Institutional Investors) ownership ratio. For each firm–year, we annualize QFII by taking the last disclosure within the calendar year (baseline). The main denominator is free-float A-shares (QFII_A); as a robustness check, we recompute the ratio using total shares (QFII_T). Additionally, firm–years within the listing window that have no QFII record are coded as zero (true absence).

From the Huazheng ESG Database, we collect firm-level ESG ratings for 2009–2022 at an annual frequency, including the composite score and the environmental (E), social (S), and governance (G) pillars. In the baseline, the composite ESG score serves as the moderator; in extensions, we replace the moderator with E, S, and G, respectively, to compare which pillar drives the moderating effect (Fatemi et al., 2018). From CSMAR, we also compile the following annual control variables: firm size (FirmSize), firm age (FirmAge), Leverage, ROA, ROE, board size (LnBoard), proportion of independent directors (IndeptDirect), duality, and proportion of the board’s shareholdings (BoardHld).

After removing observations with missing key fields, we further winsorized continuous variables at the 1% tails on both ends. The final sample covers the period from 2009 to 2022 and contains 4940 publicly listed Chinese firms, yielding about 40,334 firm–year observations.

3.2. Methodology

This study examines whether the board’s overseas experience (BOE) influences the shareholdings of Qualified Foreign Institutional Investors (QFIIs) in Chinese A-share listed companies. Building on the prior literature related to board characteristics (Gu et al., 2023; Khan et al., 2023; Liang et al., 2024; Liu & Kang, 2025; Miletkov et al., 2014) and foreign institutional ownership (Aghion et al., 2013; Guo & Zheng, 2021; Jin et al., 2016; Korkeamäki et al., 2019; Leuz et al., 2009), this study integrates two strands of research by examining how the board’s overseas experience affects QFII shareholdings in Chinese A-share firms. The baseline empirical model is specified as:

$$\begin{array}{l}QFI{I\_A}_{i,t}={\beta }_0+{\beta }_{1}OE{B}_{i,t}+{\beta }_{2}FirmSiz{e}_{i,t}+{\beta }_{3}FirmAg{e}_{i,t}+{\beta }_{4}Leverag{e}_{i,t}+{\beta }_{5}RO{A}_{i,t}+{\beta }_{6}RO{E}_{i,t}\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\beta }_{7}LnBoar{d}_{i,t}+{\beta }_{8}IndepDirec{t}_{i,t}+{\beta }_{9}Dualit{y}_{i,t}+{\beta }_{10}BoardHld+{\beta }_{Y}YearDum\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\beta }_{I}IndDum+{\epsilon }_{i,t}\end{array}$$

(1)

where QFII_A represents the shareholding ratio of Qualified Foreign Institutional Investors scaled by the number of free-float A-shares for firm i in year t. The key explanatory variable BOE denotes the proportion of board members with overseas educational or work experience.

Control variables include firm size (FirmSize), firm age (FirmAge), leverage, return on assets (ROA), return on equity (ROE), board size (LnBoard), proportion of independent directors (IndepDirect), board-CEO duality (Duality), and board shareholding (BoardHld). Moreover, FirmSize is measured as the natural logarithm of total assets, FirmAge is defined as the natural logarithm of the number of years since listing [ln (listing year − year of IPO)], and LnBoard represents the natural logarithm of the number of board members. Taking the logarithm for these three variables helps to reduce scale differences across firms and mitigate heteroskedasticity in regression estimation (Y. Li et al., 2018). The last four governance-related controls—LnBoard, IndepDirect, Duality, and BoardHld—are ad hoc variables selected based on the focus of our study on board-level characteristics. Including these board-level ad hoc controls can improve model consistency and help isolate the specific effect of board’s overseas experience from other general board governance attributes (Miletkov et al., 2014). These control variables capture both firm characteristics and corporate governance factors that may affect foreign investors’ preferences (Liang et al., 2024).

To address unobserved heterogeneity, we incorporate fixed effects for both industry and year, denoted as IndDum and YearDum. Industry fixed effects absorb sectoral differences, while year fixed effects account for macroeconomic and regulatory shocks affecting all firms simultaneously (Leuz et al., 2009). The model is estimated using ordinary least squares (OLS), and standard errors are adjusted for heteroskedasticity and clustered at the firm level to mitigate serial correlation issues (Jin et al., 2016). All continuous variables are winsorized at the 1st and 99th percentiles to reduce the influence of outliers.

In line with standard empirical practice, we estimate four specifications to assess the robustness of the results: (1) estimation without fixed effects, (2) estimation with year fixed effects only, (3) estimation with industry fixed effects only, and (4) estimation with both year and industry fixed effects. In later robustness tests, we further substitute QFII_T (QFII ownership scaled by total shares) for QFII_A and replace BOE with the high-overseas dummy (BOE Dummy), which is constructed by first computing the median value of BOE across all firms in each year and then assigning a value of 1 if a firm’s BOE in the following year exceeds that annual median, and zero otherwise. Furthermore, we also incorporate ESG performance moderating effects in subsequent sections. Specifically, we use the composite ESG score as well as its three subdimensions—environmental (E), social (S), and governance (G)—to examine whether the moderating effect differs across the three dimensions. To ensure comparability across firms and years, all ESG indicators are standardized to a 0–1 scale by subtracting the minimum value and dividing by the range (i.e., maximum minus minimum; see J. Li & Liu, 2025).

4. Results and Discussion

4.1. Summary Statistics

Table 1 presents the descriptive statistics for the sample of Chinese A-share listed firms from 2009 to 2022, yielding 40,334 firm–year observations. The mean value of QFII_A is 0.174% with a median of 0 and a maximum of 4.654%, indicating that while some firms have a non-trivial QFII participation, the majority of firm–years are not held by QFII, consistent with the selective nature of foreign institutional investments in China. Additionally, the mean value of the key independent variable, BOE, is 0.104, indicating that on average, 10.4% of board members possess overseas experience. The standard deviation of 0.13 further suggests substantial variation across firms in the proportion of directors with overseas experience. The dummy variable OverseasHigh equals one for 62.6% of the sample, showing that more than half of firms have above-median board internationalization.

Table 1. Descriptive statistics.

	Obs	Mean	Std. Dev	Min	P25	Median	P75	Max
BOE	40,334	0.104	0.130	0	0	0.091	0.143	0.571
OverseasHigh	40,334	0.626	0.484	0	0	1	1	1
QFII_A	40,334	0.174	0.662	0	0	0	0	4.654
QFII_T	40,334	0.112	0.425	0	0	0	0	2.890
FirmSize	40,334	22.196	1.413	19.568	21.208	21.963	22.918	27.204
FirmAge	40,334	2.041	0.929	0	1.386	2.197	2.833	3.367
Leverage	40,334	0.429	0.216	0.051	0.255	0.418	0.586	0.967
ROA	40,334	0.035	0.067	−0.311	0.013	0.037	0.067	0.206
ROE	40,334	0.049	0.167	−1.045	0.028	0.069	0.113	0.346
LnBoard	40,334	2.309	0.403	1.609	2.079	2.197	2.485	3.584
IndeptDirect	40,334	0.376	0.053	0.333	0.333	0.364	0.429	0.571
Duality	40,334	0.038	0.192	0	0	0	0	1
BoardHld	40,334	0.134	0.193	0	0	0.005	0.253	0.674
Std_ESG	40,334	0.646	0.095	0.344	0.593	0.653	0.71	0.844
Std_E	40,334	0.470	0.113	0.247	0.390	0.466	0.544	0.772
Std_S	40,334	0.740	0.097	0.452	0.680	0.747	0.803	1
Std_G	40,334	0.757	0.090	0.408	0.723	0.776	0.813	0.912
Non_Mainland_BP	34,877	0.001	0.010	0	0	0	0	0.091

Note: Table 1 reports the descriptive statistics for all variables used in the empirical analyses. The dataset comprises 40,334 firm–year observations, covering A-share listed firms from 2009 to 2022.

The firm-level controls also exhibit reasonable dispersion. The mean of FirmSize (natural logarithm of total assets) is 22.196, and FirmAge averages 7 years (the exponential of 2.041). The mean leverage ratio is 0.429, while ROA and ROE average 0.035 and 0.049, respectively. Regarding board characteristics, LnBoard averages 2.309 (approximately 10 director members in raw value), the proportion of independent directors is 0.376, the CEO duality dummy equals one in 3.8% of observations, and the average board shareholding is 0.134. These statistics align with the governance structure of typical A-share non-financial firms. Standardized ESG scores are also reported, with mean Std_ESG, Std_E, Std_S, and Std_G values of 0.646, 0.470, 0.740, and 0.757, respectively.

Table 2 reports the Pearson correlation matrix. BOE is positively correlated with both QFII_A (0.078 ***) and QFII_T (0.069 ***). The ESG composite and its subdimensions are also positively related to QFII measures, suggesting that firms with higher sustainability ratings tend to attract more foreign institutional attention. Among control variables, the correlation patterns are economically intuitive: larger firms tend to be older (0.352 *** for FirmSize & FirmAge), and leverage is positively correlated with FirmSize (0.515 ***). Most correlations among explanatory variables are below 0.55 in absolute value, implying no severe multicollinearity concerns (Aghion et al., 2013; Guo & Zheng, 2021).

Table 2. Pearson correlation coefficients between regression variables.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)
(1) BOE	1
(2) OverseasHigh	0.588 ***	1
(3) QFII_A	0.078 ***	0.033 ***	1
(4) QFII_T	0.069 ***	0.030 ***	0.929 ***	1
(5) FirmSize	0.120 ***	−0.014 ***	0.084 ***	0.099 ***	1
(6) FirmAge	−0.063 ***	−0.054 ***	−0.050 ***	0.023 ***	0.352 ***	1
(7) Leverage	−0.030 ***	−0.001	−0.005	0.016 ***	0.515 ***	0.384 ***	1
(8) ROA	0.030 ***	0.039 ***	0.085 ***	0.071 ***	−0.011 **	−0.216 ***	−0.350 ***	1
(9) ROE	0.031 ***	0.042 ***	0.075 ***	0.067 ***	0.094 ***	−0.130 ***	−0.216 ***	0.861 ***	1
(10) LnBoard	0.028 ***	−0.017 ***	0.076 ***	0.035 ***	0.160 ***	−0.397 ***	−0.011 **	0.119 ***	0.112 ***	1
(11) IndeptDirect	0.052 ***	−0.041 ***	0.011 **	0.009 *	−0.005	−0.012 **	−0.017 ***	−0.020 ***	−0.021 ***	−0.273 ***	1
(12) Duality	0.012 **	−0.004	0	−0.003	−0.068 ***	−0.081 ***	−0.057 ***	0.009 *	0	−0.025 ***	0.062 ***	1
(13) BoardHld	0.029 ***	−0.021 ***	0.007	−0.027 ***	−0.328 ***	−0.541 ***	−0.340 ***	0.148 ***	0.070 ***	0.034 ***	0.077 ***	0.095 ***	1
(14) Std_ESG	0.070 ***	0.034 ***	0.081 ***	0.075 ***	0.237 ***	−0.153 ***	−0.088 ***	0.239 ***	0.238 ***	0.184 ***	0.057 ***	−0.002	0.085 ***	1
(15) Std_E	0.043 ***	−0.024 ***	0.041 ***	0.045 ***	0.262 ***	0	0.099 ***	0.016 ***	0.033 ***	0.118 ***	−0.011 **	−0.020 ***	−0.022 ***	0.535 ***	1
(16) Std_S	0.063 ***	−0.042 ***	0.044 ***	0.048 ***	0.203 ***	−0.040 ***	0.059 ***	0.090 ***	0.102 ***	0.069 ***	0.006	0.008 *	0.086 ***	0.654 ***	0.272 ***	1
(17) Std_G	0.027 ***	0.089 ***	0.072 ***	0.055 ***	0.037 ***	−0.223 ***	−0.258 ***	0.294 ***	0.267 ***	0.173 ***	0.093 ***	0.007	0.089 ***	0.686 ***	0.096 ***	0.038 ***	1

Note: This table reports Pearson correlation coefficients among the core independent variable (BOE), the dependent variable (QFII_A), all control variables, the moderating variables (ESG and its subdimensions), and additional variables used in robustness analyses. The dataset consists of 40,334 firm–year observations, encompassing approximately 4940 publicly listed Chinese firms over the period from 2009 to 2022. *, **, and *** refer to significance at the 10%, 5%, and 1% levels, respectively.

4.2. Baseline Regression Results

Our study examines the relationship between board’s overseas experience and foreign investors’ holdings in Chinese A-share listed firms. Our main variable of interest is BOE, measured as the proportion of directors with overseas educational or professional backgrounds. The dependent variable QFII_A represents the number of shares held by Qualified Foreign Institutional Investors scaled by free-float A-shares of a firm. All regressions are estimated using ordinary least squares (OLS), and standard errors are adjusted for heteroskedasticity and within-firm serial correlation by clustering at the firm level (Jin et al., 2016). Table 3 reports the results across four model specifications. Following prior studies on foreign ownership and board characteristics (Aghion et al., 2013; Miletkov et al., 2014; Jin et al., 2016; Guo & Zheng, 2021; Liang et al., 2024), we control for firm-level factors (FirmSize, FirmAge, Leverage, ROA, ROE) and board-level variables (including LnBoard, IndepDirect, Duality, and BoardHld) that may influence foreign investors’ decisions. We include year fixed effects and industry fixed effects to control for unobservable macroeconomic and sectoral shocks (Leuz et al., 2009).

Table 3. Baseline regression of QFII_A on overseas experience of board members (BOE).

Variable	(1)	(2)	(3)	(4)
	QFII_A	QFII_A	QFII_A	QFII_A
BOE	0.303 ***	0.306 ***	0.234 ***	0.239 ***
	(0.056)	(0.057)	(0.055)	(0.055)
FirmSize	0.044 ***	0.046 ***	0.039 ***	0.041 ***
	(0.008)	(0.009)	(0.006)	(0.006)
FirmAge	−0.035 ***	−0.035 ***	−0.027 ***	−0.027 ***
	(0.009)	(0.009)	(0.009)	(0.009)
Leverage	−0.027	−0.039	−0.037	−0.050 *
	(0.032)	(0.031)	(0.029)	(0.029)
ROA	0.834 ***	0.797 ***	0.952 ***	0.918 ***
	(0.171)	(0.173)	(0.131)	(0.132)
ROE	−0.077	−0.065	−0.126 ***	−0.116 ***
	(0.058)	(0.056)	(0.040)	(0.039)
LnBoard	0.061 ***	0.054 **	0.042 **	0.034 *
	(0.022)	(0.023)	(0.019)	(0.019)
IndeptDirect	0.241 **	0.223 **	0.186 *	0.169
	(0.107)	(0.108)	(0.105)	(0.106)
Duality	0.003	0.002	−0.000	−0.001
	(0.018)	(0.018)	(0.018)	(0.018)
BoardHld	−0.027	−0.016	−0.039	−0.028
	(0.034)	(0.035)	(0.034)	(0.035)
Industry FE	No	No	Yes	Yes
Year FE	No	Yes	No	Yes
Observations	40,334	40,334	40,334	40,334
Adjusted R2	0.023	0.028	0.048	0.053

Note: This table reports the regression results for the baseline models examining the effect of directors’ overseas experience (BOE) on foreign investors’ holdings (QFII_A). Absolute t-statistics, based on heteroskedasticity-consistent standard errors clustered at the firm level, are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

Table 3 presents the regression results of QFII_A on BOE and the control variables across four model specifications. The coefficient on BOE is positive and statistically significant at the 1% level in all columns, confirming that firms with higher proportions of directors possessing overseas experience attract greater foreign institutional ownership. Specifically, in the full model with both industry and year fixed effects, the estimated coefficient on BOE is 0.239. Everything else being equal, a one standard deviation increase in BOE (0.13) leads to a 0.0311 (0.239 × 0.13 = 0.0311) increase in QFII_A, representing a 17.87% (0.0311/0.174 = 0.1787) rise relative to the sample mean of QFII_A (0.174). This magnitude indicates that the effect of board’s overseas experience on foreign investors’ holdings is economically significant, suggesting a meaningful improvement in foreign investors’ participation when board internationalization increases. This finding suggests that directors’ overseas experience has a positive effect on foreign investors’ holdings, consistent with the view that overseas experience facilitates the transfer of governance practices, knowledge, and skills valued by global investors and serves as a credible signal that attracts foreign capital.

Control variables generally show the expected signs. FirmSize is positive and strongly significant across all models, implying that larger firms are more attractive to overseas investors due to their visibility, liquidity, and information availability (Dahlquist & Robertsson, 2001). FirmAge carries a negative and significant coefficient, suggesting that younger firms, possibly characterized by higher growth potential, appeal more to foreign institutions (Aggarwal et al., 2005). ROA is positive and highly significant, while ROE becomes negative when both fixed effects are included, possibly reflecting the leverage–profitability interaction once asset-based returns are controlled. Leverage is weakly negative in the most saturated specification, which aligns with the idea that foreign investors gravitate toward firms exhibiting lower levels of financial risk (Miletkov et al., 2014; Khan et al., 2023). LnBoard and IndepDirect both exhibit positive coefficients, indicating that larger and more independent boards facilitate foreign participation, although the latter loses significance when full fixed effects are applied. Duality and BoardHld are insignificant, suggesting limited influence on foreign ownership in this context.

4.3. Moderating Effect of ESG Performance

In this section, we investigate whether a firm’s ESG performance moderates the relationship between directors’ overseas experience (BOE) and foreign investors’ holdings (QFII_A).

Our prior analysis suggests that overseas-experienced directors can strengthen corporate governance, support strategic initiatives that improve firm performance, and mitigate the information asymmetry faced by foreign investors—factors that collectively attract foreign capital. We therefore expect the positive effect of overseas board experience to be stronger when these directors can play an active and effective monitoring and advising role and when controlling shareholders and management are receptive to their input. To proxy for the effectiveness of such monitoring and advising, we use the firm’s ESG performance.

This proxy is motivated by the fact that ESG has a longer and more established history in many developed markets. Directors who studied or worked abroad are more likely to view ESG as integral to firm strategy, risk management, and stakeholder relations, and to recognize its relevance for firm value and stakeholder interests (Wen et al., 2020). Accordingly, among firms with overseas-experienced directors, stronger ESG performance may reflect greater director influence on corporate policies, a more effective monitoring/advising role, and a stronger long-term orientation (Zhang et al., 2018).

As noted previously, to ensure comparability across firms and years, we standardize all ESG indicators including E, S, and G individually to a 0–1 scale using the min-max method (Friede et al., 2015). Specifically, we subtract the minimum value and divide by the range (maximum minus minimum). In the following calculation, we use a standardized ESG/E/S/G score, quoted as ESG, E, S, and G individually.

$$ESG/E/S{/G}_{i,t}={\displaystyle \frac{ESG/E/S/G Scor{e}_{i,t}-\mathit{min}\left(ESG/E/S/G Scor{e}_{i,t}\right)}{\mathit{max}\left(ESG/E/S/G Scor{e}_{i,t}\right)-min(ESG/E/S/G Scor{e}_{i,t})}}$$

(2)

Specifically, we include the standardized ESG score (ESG) and its interaction term with BOE in the baseline model to test this moderating effect.

$$\begin{array}{l}QFI{I\_A}_{i,t}={\beta }_0+{\beta }_{1}OE{B}_{i,t}+{\beta }_{2}ES{G}_{i,t}+{\beta }_{3}OE{B}_{i,t}\times ES{G}_{i,t}+{\beta }_{4}FirmSiz{e}_{i,t}+{\beta }_{5}FirmAg{e}_{i,t}\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\beta }_{6}Leverag{e}_{i,t}+{\beta }_{7}RO{A}_{i,t}+{\beta }_{8}RO{E}_{i,t}+{\beta }_{9}LnBoar{d}_{i,t}+{\beta }_{10}IndepDirec{t}_{i,t}\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} +{\beta }_{11}Dualit{y}_{i,t}+{\beta }_{12}BoardHld+{\beta }_{Y}YearDum++{\beta }_{I}IndDum+{\epsilon }_{i,t}\end{array}$$

(3)

The results are reported in Table 4. In column (1), when ESG performance and its interaction term with BOE are included, the coefficient on BOE becomes −0.353 (t = 1.09, insignificant), while the interaction term (BOE × ESG) is 0.906 and significant at the 10% level (t = 1.78). In previous section, Table 3 reports that the coefficient on BOE is 0.239 and significant at the 1% level, confirming a positive association between board’s overseas experience and QFII holdings in the baseline model. However, the current result in Table 4 will not overturn the result in Table 3, and this moderating effect pattern indicates that ESG performance strengthens the positive effect of board’s overseas experience on foreign institutional ownership. The change in the BOE coefficient does not imply a reversal of the main relationship; rather, it reflects that the coefficient on BOE in the interaction model captures the conditional effect when ESG = 0. The overall marginal effect of BOE on QFII_A can be expressed as

$${\displaystyle \frac{\partial QFII\_A}{\partial OEB}}={\beta }_1+{\beta }_{3}ESG$$

(4)

Table 4. The moderating effect of ESG/E/S/G performance.

Variable	(1)	(2)	(3)	(4)
	QFII_A	QFII_A	QFII_A	QFII_A
	(ESG)	(E)	(S)	(G)
BOE	−0.353	0.456 **	−0.717 *	−0.337
	(1.09)	(2.13)	(1.94)	(1.05)
ESG/E/S/G	0.090	0.061	0.009	0.108 *
	(1.48)	(1.05)	(0.14)	(1.90)
Interaction Term	0.906 *	−0.458	1.280 **	0.758 *
	(1.78)	(1.05)	(2.56)	(1.70)
FirmSize	0.035 ***	0.040 ***	0.038 ***	0.038 ***
	(5.33)	(6.26)	(5.74)	(5.98)
FirmAge	−0.025 ***	−0.026 ***	−0.026 ***	−0.025 ***
	(2.61)	(2.80)	(2.70)	(2.69)
Leverage	−0.034	−0.050 *	−0.051 *	−0.027
	(1.17)	(1.69)	(1.72)	(0.89)
ROA	0.890 ***	0.913 ***	0.905 ***	0.882 ***
	(6.73)	(6.96)	(6.86)	(6.70)
ROE	−0.119 ***	−0.114 ***	−0.118 ***	−0.117 ***
	(3.02)	(2.90)	(2.98)	(2.98)
LnBoard	0.030	0.034 *	0.035 *	0.029
	(1.62)	(1.83)	(1.87)	(1.54)
IndeptDirect	0.141	0.170	0.174 *	0.121
	(1.34)	(1.61)	(1.65)	(1.14)
Duality	−0.001	−0.001	−0.001	−0.001
	(0.03)	(0.05)	(0.06)	(0.05)
BoardHld	−0.032	−0.028	−0.034	−0.028
	(0.93)	(0.81)	(1.00)	(0.82)
Industry FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Observations	40,334	40,334	40,334	40,334
Adjusted R2	0.056	0.056	0.056	0.056

Note: This table provides the results of the regressions examining the moderating effect of ESG performance, including its environmental (E), social (S), and governance (G) components. Absolute t-statistics, based on heteroskedasticity-consistent standard errors clustered at the firm level, are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

Using the estimated coefficients, the marginal effect becomes positive when standardized ESG score ≥ 0.39 (0.353/0.906). This finding suggests that for firms with moderate to high ESG performance, the positive effect of directors’ overseas experience on foreign ownership becomes stronger, as ESG disclosure mitigates information asymmetry and enhances investors’ confidence in governance quality (Fatemi et al., 2018).

Columns (2) to (4) further decompose ESG into environmental (E), social (S), and governance (G). We also standardize each pillar like Equation (2). The coefficient on the interaction term BOE × Social is 1.280 (t = 2.56) and statistically significant at the 5% level, while that on BOE × Governance is 0.758 (t = 1.70) and significant at the 10% level. In contrast, the coefficient on BOE × Environmental is −0.458 (t = 1.05) and statistically insignificant. These results imply that the moderating role of ESG performance is mainly driven by the social and governance dimensions. Unlike environmental outcomes, which are often driven by external regulations or technical operations, S and G performance directly reflects the board’s discretionary influence over internal controls, shareholder rights, and stakeholder engagement. Accordingly, S and G provide a more targeted proxy for the effectiveness of overseas-experienced directors’ monitoring and advising.

The model incorporates industry and year fixed effects, and the adjusted R² of 0.056 is close to that in the baseline regression (0.053) in Table 3, confirming the stability and explanatory consistency of the model. Overall, the results demonstrate that ESG performance—particularly its social and governance components—enhances the positive association between board’s overseas experience and foreign investors’ holdings.

4.4. Robustness Tests: Alternative Measure of Foreign Investors’ Holdings

To check whether the main conclusion depends on the particular measure of foreign institutional ownership, we replace the dependent variable—QFII_A (QFII shareholdings relative to A-shares in free float)—by QFII_T, defined as QFII shareholdings relative to total shares outstanding and rerun the regressions. The distinction is important because free float excludes closely held shares, such as restricted shares and long-term holdings by insiders and strategic investors, which are not intended for active trading. As a result, QFII_A may mechanically inflate ownership ratios in firms with a smaller tradable base. Using total shares outstanding mitigates this scaling concern and provides a more comprehensive ownership measure. Moreover, this alternative specification helps assess whether foreign investors concentrate on short-term trading within the tradable float or instead hold economically meaningful positions relative to the firm’s entire equity base, consistent with longer-term investment interests.

As shown in Table 5, the coefficient on BOE under the baseline measure in column (1) is 0.239 (t = 4.38, 1% level). When the dependent variable is replaced by QFII_T in column (2), the coefficient (0.142, t = 3.95) remains positive and statistically significant at the 1% level. The magnitude is slightly smaller, which is expected because QFII_T, by definition, accounts for total shares, yet the direction and significance of the effect persist. The adjusted R² (0.051) is nearly identical to that of the baseline regression (0.053), confirming model stability. Overall, the evidence indicates that using different definitions of the dependent variable does not alter the positive link found between board overseas experience and the level of foreign ownership, which is robust to alternative definitions of the dependent variable.

Table 5. Alternative measure of QFII relative to total shares outstanding.

Variable	(1)	(2)
	QFII_A	QFII_T
BOE	0.239 ***	0.142 ***
	(4.38)	(3.95)
FirmSize	0.041 ***	0.025 ***
	(6.47)	(5.63)
FirmAge	−0.027 ***	0.018 ***
	(2.86)	(2.99)
Leverage	−0.050 *	−0.030
	(1.69)	(1.47)
ROA	0.918 ***	0.566 ***
	(6.95)	(6.37)
ROE	−0.116 ***	−0.059 **
	(2.93)	(2.17)
LnBoard	0.034 *	0.012
	(1.82)	(0.99)
IndeptDirect	0.169	0.089
	(1.60)	(1.28)
Duality	−0.001	0.004
	(0.06)	(0.37)
BoardHld	−0.028	−0.008
	(0.80)	(0.37)
Industry FE	Yes	Yes
Year FE	Yes	Yes
Observations	40,334	40,334
Adjusted R2	0.056	0.051

Note: This table reports a robustness check using QFII_T as an alternative measure of foreign institutional ownership. Absolute t-statistics, based on heteroskedasticity-consistent standard errors clustered at the firm level, are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

4.5. Robustness Tests: Alternative Measures of Board’s Overseas Experience

To confirm that the positive relationship between directors’ overseas experience and foreign investors’ holdings is not sensitive to the way BOE is measured in the baseline estimations, we employ two alternative proxies for BOE as robustness checks. First, we construct a high-overseas dummy variable (BOEDummy). This measure is constructed by first computing the median value of BOE across all firms in each year and then assigning a value of one if a firm’s BOE in the following year exceeds that annual median, and zero otherwise. This dummy variable captures whether firms possess a relatively high proportion of directors with overseas experience compared to their peers at the same year. Second, we use the natural logarithm of BOE plus one (LnBOE) to address the skewness and zero values of the BOE ratio, providing a continuous and scale-adjusted representation of the BOE ratio (Ferreira & Matos, 2008).

As reported in Table 6, the baseline specification in column (1) shows a significantly positive coefficient on BOE (0.239, t = 4.38, p < 0.01). When BOE is replaced by BOEDummy in column (2), the coefficient remains positive and significant at the 1% level (0.036, t = 3.20), suggesting that corporations with a larger share of overseas-experienced directors continue to attract greater foreign institutional ownership. Similarly, when using LnBOE in column (3), the coefficient remains positive and significant at the 1% level (0.277, t = 4.45). These results confirm that the main finding—that directors’ overseas experience enhances foreign institutional ownership—does not depend on the specific measurement of BOE. Overall, the results demonstrate that the positive relationship between board’s overseas experience and foreign investors’ holdings is robust to alternative constructions of the key independent variable.

Table 6. Alternative measures of overseas experience of board members (BOE).

Variable	(1)	(2)	(3)
	QFII_A	QFII_A	QFII_A
BOE	0.239 ***
	(4.38)
BOEDummy		0.036 ***
		(3.20)
LnBOE			0.277 ***
			(4.45)
FirmSize	0.041 ***	0.044 ***	0.041 ***
	(6.47)	(6.89)	(6.47)
FirmAge	−0.027 ***	−0.030 ***	−0.027 ***
	(2.86)	(3.14)	(2.87)
Leverage	−0.050 *	−0.058 *	−0.050 *
	(1.69)	(1.96)	(1.69)
ROA	0.918 ***	0.918 ***	0.918 ***
	(6.95)	(6.93)	(6.95)
ROE	−0.116 ***	−0.117 ***	−0.116 ***
	(2.93)	(2.95)	(2.93)
LnBoard	0.034 *	0.033 *	0.034 *
	(1.82)	(1.75)	(1.80)
IndeptDirect	0.169	0.195 *	0.171
	(1.60)	(1.84)	(1.62)
Duality	−0.001	−0.001	−0.001
	(0.06)	(0.05)	(0.06)
BoardHld	−0.028	−0.031	−0.028
	(0.80)	(0.88)	(0.82)
Industry FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Observations	40,334	40,334	40,334
Adjusted R2	0.056	0.054	0.055

Note: This table reports robustness results using alternative proxies for directors’ overseas experience, including BOEDummy and LnBOE. Absolute t-statistics, based on heteroskedasticity-consistent standard errors clustered at the firm level, are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

4.6. Robustness Tests: Subsample Analyses Excluding Crisis Periods

To rule out the concern that the baseline relation is driven by turbulent episodes, we re-estimate the model on subsamples that exclude major shock windows in China’s A-share market. Specifically, column (2) removes the Global Financial Crisis (2008–2009). Although the crisis originated in advanced economies, China was relatively less affected and was at times perceived by foreign investors as an undervalued or relatively resilient market during that period. Excluding these years helps ensure that our results are not driven by crisis-specific capital reallocation or valuation effects. Column (3) excludes the COVID-19 period (2020–2022), which was characterized by heightened uncertainty, mobility restrictions, and market-wide disruptions that may have influenced both board composition and foreign investment decisions. This approach follows standard robustness practice by keeping the specification unchanged and varying the sample window.

As reported in Table 7, the core effect remains intact. In the full sample in column (1), the coefficient on BOE is 0.239 (t = 4.38). After excluding 2008–2009 in column (2), the coefficient on BOE becomes 0.243 (t = 4.46); after excluding 2020–2022 in column (3), the coefficient on BOE (0.274, with t = 3.90) remains similar in magnitude to that in our baseline estimation. The magnitudes are slightly larger in both restricted windows, indicating that the positive association between board’s overseas experience and foreign investors’ holdings is not an artifact of crisis-era behavior. Model fit is comparable across windows since the adjusted R² moves from 0.056 (full sample) to 0.053 (excluding 2008–2009) and 0.081 (excluding 2020–2022). Overall, the subsample analyses confirm that our baseline finding is stable outside crisis periods.

Table 7. Robustness tests excluding Global Financial Crisis (2008–2009) and COVID-19 period (2020–2022).

Variable	(1)	(2)	(3)
	QFII_A	QFII_A	QFII_A
	(Full Sample)	(Excluding 2008–2009)	(Excluding 2020–2022)
BOE	0.239 ***	0.243 ***	0.274 ***
	(4.38)	(4.46)	(3.90)
FirmSize	0.041 ***	0.039 ***	0.055 ***
	(6.47)	(6.18)	(7.39)
FirmAge	−0.027 ***	−0.029 ***	0.001
	(2.86)	(2.98)	(0.08)
Leverage	−0.050 *	−0.045	−0.057
	(1.69)	(1.51)	(1.63)
ROA	0.918 ***	0.902 ***	1.395 ***
	(6.95)	(6.81)	(8.04)
ROE	−0.116 ***	−0.113 ***	−0.232 ***
	(2.93)	(2.84)	(4.33)
LnBoard	0.034 *	0.032 *	0.013
	(1.82)	(1.73)	(0.60)
IndeptDirect	0.169	0.173	0.159
	(1.60)	(1.63)	(1.32)
Duality	−0.001	−0.003	0.018
	(0.06)	(0.17)	(0.81)
BoardHld	−0.028	−0.030	−0.066 *
	(0.80)	(0.85)	(1.70)
Industry FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Observations	40,334	38,953	27,405
Adjusted R2	0.056	0.053	0.081

Note: This table reports robustness checks by re-estimating the baseline model after excluding the Global Financial Crisis (2008–2009) and the COVID-19 period (2020–2022). Absolute t-statistics, based on heteroskedasticity-consistent standard errors clustered at the firm level, are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

4.7. Endogeneity and Instrumental Variable Analysis

Endogeneity may arise in Equation (1) if unobservable firm-level attributes or reverse causality jointly affect board’s overseas experience (BOE) and foreign investors’ holdings (QFII_A). To mitigate this concern, we implement a two-stage least squares approach. Following the related literature (e.g., Lai et al., 2020; Lei et al., 2025), we construct our instrumental variable as the one-year lagged ratio of board members whose birthplace is outside mainland China, denoted as Non_Mainland_BP. Directors born outside mainland China (including Hong Kong, Taiwan, Macau, and foreign countries) are more likely to possess overseas education or work experience due to lower language and cultural barriers.

Importantly, the birthplace of directors is plausibly exogenous to a firm’s ownership structure, as it is predetermined and unlikely to be influenced by contemporaneous firm-level decisions. To further alleviate the possibility of reverse causality—that is, the likelihood that firms with greater overseas ownership appoint non-mainland-born directors—we use the one-year lagged value of the instrument. All 2SLS specifications retain industry and year fixed effects and cluster standard errors at the firm level, consistent with the baseline model. In the first step, BOE is regressed on the instrument variable and controls to obtain the predicted BOE. In the second step, QFII_A is regressed on this predicted BOE and the same set of controls.

Table 8 reports the results from the 2SLS regression. The coefficient on Non_Mainland_BP in the first stage equals 1.737, significant at the 1% level (t = 12.24), with identification diagnostics confirming both relevance and validity. Although the Kleibergen–Paap rk Wald F statistic (9.701) suggests potential weak instrument concerns relative to conventional thresholds (i.e., 10), the Anderson–Rubin test strongly rejects the null hypothesis (χ² = 50.11, p < 0.01), indicating that our inference is robust to weak instruments. In the second stage, the estimated coefficient on predicted BOE is 0.586, significant at the 10% level (t = 1.79), demonstrating a positive and statistically significant effect of board’s overseas experience on foreign investors’ holdings. Overall, the results derived from the instrumental variable regression not only align with but reinforce the preceding OLS findings. The significantly larger coefficient obtained from the 2SLS estimation compared with the OLS model (0.586 vs. 0.239) suggests that the baseline relationship is biased downward due to endogeneity, which has now been effectively corrected. Therefore, Table 8 provides evidence consistent with the causal direction runs from board’s overseas experience to foreign investors’ holdings, rather than the inverse.

Table 8. Two-stage least squares regression results.

Variable	(1)	(2)
	BOE	QFII_A
Non_Mainland_BP	1.737 ***
	(12.24)
Predicted BOE		0.586 *
		(1.79)
FirmSize	0.018 ***	0.032 ***
	(10.92)	(3.62)
FirmAge	−0.016 ***	−0.006
	(6.41)	(0.48)
Leverage	−0.046 ***	−0.018
	(5.09)	(0.54)
ROA	0.012	1.029 ***
	(0.35)	(7.38)
ROE	−0.013	−0.132 ***
	(1.26)	(3.32)
LnBoard	−0.009 *	0.017
	(1.81)	(0.78)
IndeptDirect	0.060 **	0.158
	(2.06)	(1.37)
Duality	0.003	−0.001
	(0.77)	(0.05)
BoardHld	−0.006	0.019
	(0.57)	(0.51)
Industry FE	Yes	Yes
Year FE	Yes	Yes
Observations	34,876	34,876
Adjusted R2	0.099	0.055

Note: This table presents the two-stage least squares (2SLS) estimation results using one-year lagged proportion of non-mainland born directors as the instrumental variable. Absolute t-statistics, based on heteroskedasticity-consistent standard errors clustered at the firm level, are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

5. Conclusions

Foreign investors are becoming an increasingly important source of capital in China, the world’s second-largest equity market. At the same time, the growing pool of internationally trained professionals and returning talent is reshaping managerial expertise and corporate governance practices among Chinese firms. Against this backdrop, we examine whether directors’ overseas experience is associated with foreign investors’ holdings in Chinese A-share firms.

Using a large panel of Chinese listed firms from 2009 to 2022, we document a positive association between the proportion of directors with overseas experience and foreign institutional ownership. The relation is economically meaningful and robust to alternative model specifications, variable constructions, and sample restrictions, and it remains after addressing endogeneity concerns. Additional analysis indicates that the association is stronger when overseas-experienced directors are more likely to exert influence through monitoring and advising roles. Overall, our findings suggest that directors’ overseas experience facilitates the diffusion of governance practices, knowledge, and skills valued by global investors and serves as a credible signal that helps attract foreign capital.

These findings should be interpreted in light of several limitations. First, our analysis focuses on Chinese A-share firms, and China’s institutional and regulatory environment may limit the generalizability of the results to other settings. Second, our measures of directors’ overseas experience (and related constructs such as ESG performance) rely on existing database classifications, which may not fully capture heterogeneity in the quality, duration, and relevance of overseas experience. Future research could extend our framework by unpacking such heterogeneity—for example, distinguishing between overseas education versus work experience, or testing whether experience in common-law jurisdictions (e.g., the U.S. or U.K.) provides a stronger governance signal than experience in other regions. Future work could also examine how regulatory reforms and changes in foreign investor access shape the strength of the documented relation.